Cleaning in place (CIP) of ÄKTA process™ chromatography system with a recirculation procedure

Precleaning of parts

The system was first cleaned externally with lint free towels wetted with 70% ethanol. All tanks were cleaned with a detergent solution using a dish washing brush, purified water, and rinsed with 20% ethanol.

The hoses, valves, reducers, manifolds, seals, TC-clamps, and blinds of the system were autoclaved (121°C, 20 min) before attachment. Filters (0.2 µm 10", Parker Murus PROPOR HC) were attached to the pipes of the incoming NaCl and purified water/ethanol liquids. All liquids, except for NaOH, were filtered (0.2 µm).

Precleaning of system

An autoclaved manifold was attached to the four CIP inlets. All seals and interaction points were treated with 70% ethanol. The tube distributing the cleaning agent (1 M NaOH) was attached to the manifold inlet.

The system was run manually during the cleaning of the CIP block. Each inlet was opened and flushed with 1 M NaOH for approximately 20 s at a pump speed of 50%. The system was set to bypass mode with inlet A1 and outlet 1 valves open during those flushes of the CIP valves.

After the flushes, the system was left for 1 h of static hold. A rinse of the wetted flow path followed with sterile filtered 20% ethanol. All valves on the CIP block were opened and closed a couple of times during the run with the pump speed at 75%. The whole wetted flow path was filled with 20% ethanol and remained filled until the next procedure began.

The manifold on the CIP-inlet was removed and the tubes distributing the liquids (0.22 µm sterile filtered purified water, 0.22 µm sterile filtered 20% ethanol, 0.22 µm sterile filtered 0.9% NaCl, and 1 M NaOH) were attached to positions 1 through 3 on the CIP inlet.

Before attaching the tubes, autoclaved sampling equipment for testing incoming solutions was attached. Position 4 was capped.

A UNICORN™ method was created to automate the precleaning procedure. The procedure included a rinse with filtered water, application of 1 M NaOH cleaning agent, including a static hold for 1 h, and a wash-out step using sterile filtered 20% ethanol. Liquids were applied with 25% to 75% of the maximum pump speed. A manual diaphragm valve was attached to the outlet hose to apply a back pressure of at least 1 bar, (14 psi, 0.1 MPa).

Preparation and application of challenging organism suspension

The precleaned system was primed with 0.9% NaCl before application of the challenging organism suspension. Tryptic soy agar (TSA) plates streaked with challenging organism were incubated at 37°C overnight. Fresh colonies from this plate were transferred to 200 mL of autoclaved tryptic soy broth (TSB) media and left shaking in 37°C overnight.

Based on measured OD of the preculture and the assumption that 1 OD is approximately 1.5 × 10⁸ colony-forming units (CFU)/mL for S. aureus, calculations were made on the volumes needed of the preculture to be added in approximately 20 L of filtered 0.9% NaCl solution to obtain the final concentrations of 10⁶ to 10⁸ CFU/mL for the challenging organism in the suspension.

The suspension was applied to the system with 10% of the maximum pump speed bypassing the CIP valves. The procedure was automated with a UNICORN™ method. The infected system was left for 16 to 20 h at RT before starting the cleaning procedure. An overview of the challenging organism loading is described below:

1. All A inlets
2. All B inlets
3. All C inlets
4. Air trap
5. Filter
6. All column positions
7. All outlets

Cleaning procedure

Liquids were filtered through 0.2 µm filters except for the 1 M NaOH cleaning agent. The overall steps in the method were:

1. Flush the entire wet flow path with sterile filtered purified water (50% pump speed).
2. Apply 1 M NaOH cleaning agent and recirculate the agent for 1 h, evenly distributing throughout the flow path through predetermined distribution patterns achieved by combinations of valve openings. Distribution patterns were repeated throughout the recirculation time (75% pump speed, dual pump mode).
3. Replace the cleaning agent in the flow path by first rinsing with sterile filtered purified water followed by application of sterile filtered 20% ethanol as storage solution or directly use 10 or 100 mM NaOH as both rinsing and storage solution in a combined application step (50% pump speed).

Endotoxin samples were collected. The system was drained of liquid before microbial sampling took place. A manual diaphragm valve was attached to the outlet hose to apply a back pressure of at least 1 bar (14 psi, 0.1 MPa).

Microbial sampling

Microbial samples were taken at predetermined sites after the system was drained of fluids (see the swab sampling points in Tables 2 and shown visually in Figs 1 through 8 below). Applied solutions on the system were also sampled (Table 1). Microbial sampling was performed by one of the following methods:

Test method 1 microbial air sampling

Sampling of air for airborne microorganisms was conducted with a microbial air sampler (MAS). An MAS loaded with an agar plate was positioned at a suitable measuring point. When the measuring started, a predefined volume of surrounding air passed through the machine. Microorganisms were collected on the agar surface by impaction.

Test method 2 bioburden filtration test for S. aureus

Sample solutions (minimum 50 mL) were collected in sterile tubes and then filtered through a 0.45 µm cellulose nitrate membrane filters. Filters were incubated on agar plates at 30°C to 35°C for 5 d after which the plates were inspected for CFUs.

Test method 3 swab test for S. aureus

Surface samples were taken with swabs. The swab was inserted into the tube containing the isotonic swab rinse solution and vortexed for a minimum of 20 s. The solutions including the swabs were poured into Petri dishes and mixed with 30 mL of temperature-controlled molten agar. The maximum temperature of the molten agar was 45°C. After solidification, plates were incubated at 30°C to 35°C for 5 d after which the plates were inspected for CFUs.

Test method 4 viable count test for S. aureus

Samples of challenging organism suspensions were diluted in series in 0.9% NaCl. Samples from the diluted suspensions were plated on agar plates and incubated at 30°C to 35°C for 1 to 2 d after which the plates were inspected for CFUs. The concentration of the challenge organism was determined in the sampled suspensions.

Test method 5 endotoxin analysis

The method used for analysis of endotoxin content was based on the Limulus Amoebocyte lysate kinetic chromogenic assay.

Criteria for acceptance

The following acceptance criteria were used to determine if the cleaning method was successful:

1. Concentration of viable challenging organism should be 10⁶ to 10⁸ CFU/mL in the inoculum (K), post-infection sample (L), and precleaning sample (M).
2. Post-cleaning sample (I) should contain 0 CFU/mL of the challenging organism.
3. Endotoxin sample (J) should contain ≤ 0.5 EU/mL.
4. Sampling points 1 through 24 should contain 0 CFU/unit of the challenging organism.
5. Control samples should confirm that methods, materials, and handling procedures are functional.
6. A maximum of 10% of the sampling points 1 through 24 can contain other contaminants than the challenging organism (three samples in total).

Fig 1. Swab samples were collected from Inlet A (1) and Inlet B (2) as described in Table 2.

Fig 2. Swab samples were collected from connection points for liquids form pump A and B (3), pressure sensors (4 and 5), and conductivity sensors (6 and 7) as described in Table 2.

Fig 3. Swab samples were collected from the air trap filter block (8 through 12) as described in Table 2.

Fig 4. Swab samples were collected from the air trap top valve (13) and filter unit top valve (14) as described in Table 2.

Fig 5. Swab samples were collected from the pressure sensor (15) as described in Table 2.

Fig 6. Swab samples were collected from the column valve block (16 through 18), as described in Table 2.

Fig 7. Swab samples were collected from the pressure sensor (19), conductivity sensor (20-21), and pH sensor (22) as described in Table 2.

Fig 8. Swab samples were collected from the outlet valve block (23 and 24) as described in Table 2.

Tables 1, 2, and 3 present the results from the three cleaning studies including the challenging organism suspension, sample points, liquid samples, and control samples. Sample points have been denoted with a letter or number.

Figures 1 through 8 presented in the Materials and methods section show the location of the swab sample points. Based on these results, the recommendation on how operators should set up a UNICORN™ method for the application of the 1 M NaOH cleaning agent during cleaning of the system is presented below.

Table 1. Results from liquid samples collected in studies 1, 2, and 3

Table 2. Locations for collecting swab samples 1 through 24 on ÄKTA process™ CFG 10 mm PP chromatography system

Table 3. Control samples

*The results from the air sampling with MAS100EX microbial air sampler (Method: 73-5006-93 AF) have been converted to a more statistical relevant value by using the formula Pr = N (1/N + 1/(N-1) + 1/(N-2) + … 1/(N-r+1)) (Feller, 1950). Pr equals probable statistical total, and r equals the number of CFU counted on a 90 mm Petri dish. The conversion is based upon the principle that as the number of viable particles being impinged on a given plate increases, the probability of the next particle going into an "empty hole" decreases.

^† Too numerous to count (TNTC).

Study 1

The results of the serial dilution of the liquid samples (K, L, and M) of the challenging suspension are shown in Tables 4 through 6. The viable count plates confirm the S. aureus suspension was fully present prior to cleaning. The microbial air samples showed that the microorganisms and the endotoxin levels (H and J) were < 0.5 EU/mL, meeting both acceptance criteria.

However, five swab samples (8 through 12) taken on the EPDM membranes in the air trap/filter valve block proved to be positive for the challenging organism S. aureus. An additional contaminant, from the Moraxella genus, was found in sample point 16, representing valve position 1 in the column valve block. This contaminant was likely to be a false positive. As a result, the acceptance criteria were not met for this study.

Notably, it was only in the five sampled valves in the air trap/filter valve block where the challenging organism was found. The isolated finding of this clustered contamination pointed to a problem with how the method worked in this block. It was concluded that the contact time, in combination with high flow, was too short and that the contact time should be increased. This was accomplished by increasing the time that both the air trap and filter were in-line during the recirculation step which lasted 1 h. Before study 2, we increased the contact time by 30 min in the method. The rest of the time, they were put in bypass mode so that the other flow paths and valve combinations in the block would be properly cleaned.

In the case of the false positive at sample point 16, we increased focus on aseptic techniques, such as being careful to sterilize gloves and surfaces with 70% ethanol, as well as focusing on avoiding touching dirty surfaces with the swab.

Table 4. Viable count plates from study 1 of the liquid suspensions collected for sample K

Table 5. Viable count plates from study 1 of the liquid suspensions collected for sample L

Table 6. Viable count plates from study 1 of the liquid suspensions collected for sample M

Studies 2 and 3

The results of the serial dilution of the liquid samples (K, L, and M) of the challenging suspension are shown in Tables 7 through 12. The viable count plates confirm the S. aureus suspension was fully present prior to cleaning. The microbial air samples showed that the bioburden level in the air in the lab was normal and in the same range as results from previous studies. All liquid samples A through G were free from contamination (Table 1). The post-cleaning sample (I) did not contain any microorganisms, and the endotoxin levels (H and J) were < 0.5 EU/mL, meeting both acceptance criteria.

Keeping the valves for the air trap and filter units open for an additional 30 min combined with a high flow rate (75% pump speed) appeared to be necessary and sufficient to eradicate contaminants found in this valve block. Study 3 confirmed that the updated procedure was sufficient to clean the valves effectively. Since both studies met the acceptance criteria, it can be concluded that the developed method can effectively clean the ÄKTA process™ CFG chromatography system while improving cleaning efficiency.

Table 7. Viable count plates from study 2 of the liquid suspensions collected for sample K

Table 8. Viable count plates from study 2 of the liquid suspensions collected for sample L

Table 9. Viable count plates from study 2 of the liquid suspensions collected for sample M

Table 10. Viable count plates from study 3 of the liquid suspensions collected for sample K

Table 11. Viable count plates from study 3 of the liquid suspensions collected for sample L

Table 12. Viable count plates from study 3 of the liquid suspensions collected for sample M

Results from studies 1 through 3 were compiled and analyzed focusing on the ability of the 1 M NaOH cleaning agent to inactivate contaminating bacteria in relation to the parameters: flow speed, contact time, and flow distribution.

The outcome from the three optimization studies has also been taken into account in the overall analysis. The goal was to condense the conclusions drawn from these studies into recommendations on how to set up a UNICORN™ method that efficiently cleans all configurations of ÄKTA process™ chromatography systems while reducing the volume of cleaning solution needed and keeping the carbon footprint measured in kg CO₂-eq at a minimum.

Special focus was placed on components that have been challenging to clean, including valves in the inlet, air trap/filter, column and outlet blocks, and waste tubes.

Preparation before system cleaning

Precleaning manifolds and seals should be performed on inlets and outlets for efficient cleaning of the complete wetted pathway. If the system is configured with a CIP block, a manifold should initially be used for this block as well. Attached inlet manifolds should be connected using a T-branched tube that, in turn, is connected to the outlet of the CIP block.

The manifold on the outlet side must be positioned in such a way that the tube that collects all outgoing liquids points upwards to create a liquid lock. This enhances the contact time of the cleaning agent against each valve. pH probes should be replaced by dummies and cleaned according to the manufacturer's recommendations. Column valve positions should be equipped with sterilized column connections.

Cleaning of the CIP block

1. Connect the tube that distributes the cleaning agent to the inlet of the manifold attached to the CIP block.
2. Put the system in bypass mode with inlet A1 and outlet 1 opened.
3. Flush each inlet valve of the CIP block with 1 M NaOH for 20 s with pump speed at 50% and leave the system on hold for 1 h without any flow.
4. Rinse away the cleaning agent in the CIP valve block and remaining pathway with 0.22 µm sterile filtered 20% ethanol if suitable. When it is time to continue with the system cleaning, remove the manifold. Tubes distributing the cleaning agent, the wash solution, and the storage solution should be connected to the CIP block inlets. Handling must be conducted in the best aseptic manner possible.

Recommended parameters to be used in the UNICORN™ method

The UNICORN™ method should be designed so that the surface of the whole wetted flow path is in contact with the cleaning agents as long as possible during continuous high flow rates. A diaphragm valve should be attached to the outlet hose to apply a back pressure of at least 1 bar (14 psi, 0.1 MPa). By following the recommendations below, the whole wetted surface, including the surface in unmentioned components, will have enough contact time to be efficiently cleaned.

Step 1 flush flow path with purified water

1. Flush all inlet valves separately for 0.5 to 1 min using pump speed at 50% starting with the valve position with the longest flow path first in each block descending to the one with the shortest. The same procedure should be applied for the outlet valves during this step. If the number of inlet and outlet valves differ, the one with the fewest valves can be continuously flushed in a loop-like manner.
2. Continue flushing the inlet and outlet valves using the same principle as above with all pumps set in dual mode with pump speed at 50% during the remaining part of this step.

Both the air trap and filter units should be overfilled for 0.5 to 1 min at 50% pump speed and drained for at least three times the amount of time. During the remaining time of this step, units should be interchanged between in-line and bypass both modes to optimize the contact time with the high flow speed of the whole wetted flow path.

Each column position in the column valve block should be separately flushed with pump speed at 50% for at least 1 min in the following order: column position down flow, bypass bottom, column position up flow, bypass top, bypass both.

Note: Run parameters above should be adjusted depending on how the ÄKTA process™ chromatography system is configured. The number of inlets and outlets can vary, and systems can be configured with pressure control valves (PCV), including one for the C inlet, where an ÄKTA process™ system is configured for inline dilution (ILD) with a third pump, which can be attached.

Systems that have the ILD functionality can utilize its ability to run all three pumps simultaneously. The same parameters and method design mentioned above should be practiced and implemented for those extra components in the cleaning method.

Step 2 clean with 1 M NaOH

Phase 1

Inlet valve blocks

1. Run for at least 0.5 min at 75% pump speed per inlet valve starting with the last valve position in each block ascending to the first valve. This step is executed one block at a time.
2. Run at least 6.5 min with at least 37.5% pump speed (dual mode) per inlet valve, starting with the last valve position in each block ascending to the first valve. This step is executed with all downstream PCV valves opened and all pumps run in dual mode with the pump speed set to 75%.

Note: The run parameters above should be adjusted depending on the configuration of the ÄKTA process™ chromatography system in use since each PCV valve needs a total contact time with the flow of at least 30 min. The time should be divided equally on each valve in each inlet block.

Systems can be configured with a third PCV called inlet C where its own pump can be attached. Systems that have the ILD functionality can utilize its ability to run all three pumps simultaneously.

The same parameters and method design mentioned above should be practiced and implemented for this extra unit in the cleaning method. Pump speed downstream of the PCV will be collectively 75%.

Air trap and filter units

1. Overfill the air trap at least three times consecutively with the pump speed set to 75% with a drain procedure in between.
2. Open the air trap top valve and run for at least 1.5 min with the pump speed set to 75%. This step cleans the top valve and the waste tube of the air trap.
3. If the system is equipped with filter units, fill those consecutively three times with the pump speed set to 75%, with a drain procedure in between.
4. Open filter top valves and run 1.5 min with pump speed at 75% through each valve. This step cleans the top valves and waste tubes connected to the filter units.

Note: The drain procedure of the air trap and filter units are performed through the common waste tube positioned on the air trap/filter valve block. This position is controlled with its own valve. This valve, however, can never be opened at the same time as a pump is running unless you are running the out through drain function. The flow speed through this waste tube is solely based on gravitational speed if this function is not used. This valve and connected waste tube is flushed with four-unit volumes of the air trap and four-unit volumes from each existing filter unit by first filling and then draining the units. If the ÄKTA process™ chromatography system is configured with no or only one filter unit, all missing drain volumes should be compensated by running one or two times with the out through drain function. Running parameters per missing unit with this function should be set to 1 min in combination with 75% pump speed. This will generate a pump speed of 37.5% through the valve and waste tube. All wetted pathways in the air trap/filter valve block, including bypass flow paths will be efficiently cleaned if the recommended steps are followed.

Column valve block

Each column position in the column valve block is separately flushed at 75% pump speed for at least 3 min in the following order: column position down flow, bypass bottom, column position up flow, bypass top, bypass both.

Outlet valve blocks

Run at least 3 min with pump speed at 75% per outlet valve, starting with the first valve position in the block ascending to the last valve.

Phase 2

In this phase, the cleaning agent should be recirculated for 1 h at 75% pump speed in dual pump mode.

Inlet valve blocks

Valves should be opened for 30 s to be flushed, starting with valve position 1 in each inlet block (dual mode) and continuing with the remaining valves being opened and flushed in the same manner. This sequence should be looped continuously throughout phase 2.

Air trap and filter valve block

Throughout phase 2, the flow alternates between having both units in-line for 30 s and being in bypass mode for 30 s.

Column valve block

Each column position in the column valve block should be separately flushed for 30 s in the following order throughout phase 2: column position down flow, bypass bottom, column position up flow, bypass top, and bypass both.

Outlet valve blocks

Valves are opened for 30 s to be flushed, starting with valve position 1 and continuing with the remaining valves being opened and flushed in the same manner. This sequence should be looped continuously throughout phase 2.

Step 3 Replacement of cleaning agent

1. Flush all inlet valves separately for 0.25 min at 50% pump speed starting with the valve position with the longest flow path first in each block, descending to the one with the shortest. The same procedure should be applied for the outlet valves during this step. The outlet valves should be flushed for a minimum of 0.15 min. If the number of inlet and outlet valves differs, the one with the fewest valves should be continuously flushed in a loop-like manner.
2. Continue flushing the inlet and outlet valves using the same principle as above with all pumps set in dual mode with pump speed at 50% during the remaining part of this step.

Both the air trap and filter units should be overfilled for 0.5 to 1 min using 50% pump speed and drained at least three times. During the remaining time of this step, units should be interchanged between in-line and bypass both modes to optimize contact time with the high flow speed of the whole wetted flow path.

Each column position in the column valve block should be separately flushed with pump speed at 50% for at least 0.15 min in the following order: column position down flow, bypass bottom, column position up flow, bypass top, bypass both.

Step 3 should be used once or twice in the UNICORN™ method, depending on the choice of storage solution. When choosing 10 mM NaOH or 100 mM NaOH as the storage solution, it is sufficient to use this block once because you can go from cleaning solution to storage solution in one step.

If you want to use 20% ethanol as the storage solution, the block must be duplicated. The first block then aims to rinse out the cleaning solution with purified water (PW), which is followed by the second block where the storage solution is applied.

Note: Run parameters above need to be adjusted depending on how the ÄKTA process™ chromatography system is configured. The number of inlets and outlets can vary, and systems can be configured with PCV valves, including one for the C inlet where an ILD system with its own pump can be attached.

Systems that have the ILD functionality can utilize their ability to run all three pumps simultaneously.

The same parameters and method design mentioned above should be practiced and implemented for these extra components in the cleaning method.

Cleaning solution volume reductions

Methods were run with a manual diaphragm valve attached to the outlet hose to apply a back pressure of at least 1 bar (14 psi, 0.1 MPa) measured via the post-column pressure sensor. The system pressure was higher and varied with the different cleaning solutions, purified water, 1 M NaOH, and 20% ethanol.

To evaluate the improvements of this project, the method used in a previous study, CIP of ÄKTA process™ to remove S. aureus, run on a ÄKTA process™ PP 1" chromatography system, was adapted to the ÄKTA process™ CFG PP 10 mm chromatography system used in this study.

Determining starting values of liquid volumes and carbon footprint before sustainable reduction

To be able to evaluate the new cleaning method based on sustainability and a reduced climate footprint in this project, a baseline method, adapted to the system used in the previous cleaning studies, was used (Cleaning in place of ÄKTA process™ with sodium hydroxide and CIP of ÄKTA process™ to remove S. aureus). The baseline volumes are shown in Table 13.

Table 13. Baseline cleaning method fluid volumes (method 5)

Actions which reduced the cleaning fluids and kg CO₂-eq required were (Table 14–17):

1. The pump speed during application of the purified water was reduced from 75% to 50% which resulted in a reduction of volume from the approximate 125 L baseline volume to 88 to 91 L. This was a 27% to 30% reduction in the purified water needed.
2. The cleaning step, which previously contained a longer flush step followed by a hold step of 1 h (pump speed 0%), was reworked to include a short initial flush (approximately 33 L) of the entire flow path followed by a recirculation procedure of the cleaning liquid with pump speed at 75%. This resulted in a reduction of the 1 M NaOH volume from approximately 181 L to approximately 60 L. This was an approximate 67% reduction in 1 M NaOH volume needed.
3. The rinsing and storage with sterile filtered 0.9% NaCl and sterile filtered 20% ethanol, respectively was replaced with one of the following methods:
   • Method 1: rinsing with ~ 40 L of sterile filtered PW followed by application of ~ 40 L of sterile filtered 20% ethanol for storage. A reduction of pump speed in both steps from 75% to 50%. The total reduction of all applied solutions decreased from approximately 472 L baseline volume to 228 to 231 L, which was a ~ 52% total liquid reduction. This led to a reduction of kg CO₂-eq from ~ 16.8 to ~ 10.5 (~ 38% reduction).
   • Method 2 or 3: combined rinsing and storage into one step using ~ 40 L of 10 or 100 mM NaOH with a reduction of pump speed from 75% to 50%. This reduced the volume with sterile filtered 0.9% NaCl to 0 L. The total reduction of all applied solutions decreased from approximately 472 L baseline volume to 188 to 191 L, which was a ~ 60% total liquid reduction. This led to a reduction of kg CO₂-eq from ~ 16.8 to ~ 3.3 (~ 81% reduction).
   • Method 4: combined rinsing and storage in one step using 91 to 93 L of sterile filtered 20% ethanol with a reduction of pump speed from 75% to 50%. The total reduction of all applied solutions decreased from approximately 472 L baseline volume to 239 to 244 L, which was a ~ 49% total liquid reduction. This led, however, to an increase of kg CO₂-eq from ~ 16.8 to ~ 19.7 (~ 17% increase).

Table 14. Cleaning method 1 using reduced fluid volumes and kg CO₂-eq

Table 15. Cleaning method 2 (10 mM NaOH) and 3 (100 mM NaOH) using reduced fluid volumes and kg CO₂-eq

Table 16. Cleaning method 4 using reduced fluid volumes

Table 17. Calculated differences of carbon footprint equivalents of method 1 to 3 in comparison to method 4 and 5 in kg CO₂-eq

1. Cleaning components/external surfaces of ÄKTA process™ systems
2. Cleaning-in-place of ÄKTA process™ with sodium hydroxide
3. CIP of ÄKTA process™ to remove S. aureus

CY45961